RNA velocyto scRNA-seq install

Question

[Ashley]

I am performing the RNA velocyte for my scRNA-seq data.

However, I encounter a question about installing it.

My scripts are as follows:

conda install numpy scipy cython numba matplotlib scikit-learn h5py click
pip install velocyto

config.log:

**gcc: error: unrecognized command line option '-V'**
**gcc: fatal error: no input files**
compilation terminated.
configure:4445: $? = 4
configure:4434: gcc -qversion >&5
**gcc: error: unrecognized command line option '-qversion'**
gcc: fatal error: no input files
compilation terminated.
configure:4445: $? = 4
configure:4465: checking whether the C compiler works
configure:4487: gcc -fPIC -I/home/bin/bzip2-1.0.6/include  -fPIC -L/home/bin/bzip2-1.0.6/lib conftest.c  >&5
**gcc: error: unrecognized command line option '-fPIC -L/home/bin/bzip2-1.0.6/lib'**
configure:4491: $? = 1
configure:4529: result: no
configure: failed program was:
| /* confdefs.h */
| #define PACKAGE_NAME ""
| #define PACKAGE_TARNAME ""
| #define PACKAGE_VERSION ""
| #define PACKAGE_STRING ""
| #define PACKAGE_BUGREPORT ""
| #define PACKAGE_URL ""
| /* end confdefs.h.  */
| 
| int
| main ()
| {
| 
|   ;
|   return 0;
| }
configure:4534: error: in `/home/liuzhe/Software/gcc-10.1.0/gcc-10.1.0/gcc-build-10.1.0':
configure:4536: error: C compiler cannot create executables

Any response will be greatly appreciated.

Thank you.

Answer 1

[rpolicastro]

You might want to try installing velocyto with conda directly instead of using pip.

conda install -n velocyto -c conda-forge -c bioconda velocyto.py

And then to activate the environment.

conda activate velocyto

If in your originally installation of velocyto you installed the software into the conda base environment, you want to undo those changes. First, find the revision number that comes just before you installed the velocyto dependencies.

conda list -n base --revisions

Then, roll-back to the previous revision once you have the appropriate revision number.

conda install -n base --revision [revision]

As a side note, you can use mamba to greatly speed up conda.

[Friederike]

seconding this, you want to stay away from pip as much as possible when you're using conda

[ATpoint]

Not sure whether this is relevant here but on Mac I always use

echo 'auto_activate_base: false' >> ~/.condarc 

to avoid activation of the conda base environment upon terminal startup. This was to allow brew and conda to peacefully co-exist, maybe this also works with pip?

[rpolicastro]

That's a great way to avoid accidentally installing software into base. Both within conda, and outside (since pip and the like will install to base instead of your system).

For the OP the problem was that they downloaded the dependencies from conda, and then tried to install the main velocyto software using pip. For a conda environment you want to try to stay within the conda universe whenever possible, because it will ensure that software versions are compatible, and will also grab things such as system libraries that you need.